Phishing Scams Exploiting AI Technology: Domestic Phishing Case Studies 2025

Phishing Scams Exploiting AI Technology

Phishing Scams Exploiting AI Technology on the Rise

As AI technology advances, phishing scams that exploit deep fakes are on the rise. In particular, there have been reports of phishers impersonating corporate executives and giving fraudulent instructions to employees via videoconference in order to defraud them of their funds. The following are examples from outside Japan, but we should be on the lookout for similar cases in Japan in the near future.

For example, in January 2024, a Hong Kong-based company used a deep-fake video pretending to be the CFO (Chief Financial Officer) of the head office to instruct employees of the Hong Kong branch to transfer approximately HK$200 million (approximately 3.75 billion yen) via video conference.

In this scheme, the attacker elaborately mimicked the face and voice of the CFO and ordered the employees to make an urgent transfer of funds. The employees believed they were talking directly to the CFO via videoconference and followed the instructions to transfer the funds.

Another incident was reported in May 2024, in which a finance staff member was defrauded of $25 million at a UK company via videoconference using a deepfake video that impersonated the CFO.

In this case, too, the attacker used AI technology to recreate the appearance and voice of a senior executive and instructed the employee to make an urgent money transfer.

These cases are problematic because deep faking technology has become more sophisticated, making it more difficult for victims to recognize the scam compared to traditional phishing scams. In particular, there has been an increase in the use of fraudulent methods that exploit legitimate videoconferencing systems to eliminate the recipient's suspicions and successfully perpetrate the fraud.

With the evolution of AI technology, scam tactics are also becoming more sophisticated. By constantly gathering the latest information and taking appropriate countermeasures, it is possible to prevent damage before it occurs.